HoopMind
An AI Agent-based NBA intelligent question-answering demo project that supports querying key plays by round, comparing player statistics, and providing educational content about basketball stars. The system is developed through multi-model collaboration using Claude Code and DeepSeek.
Timeline (last 90 days)
The vast majority of coding for the HoopMind (NBA Chat) project was completed by Claude Code as the primary agent, supplemented by DeepSeek V4 Flash as a sub-agent.
The HoopMind project adopts checklist-driven development, first producing an RFC requirements document with clearly defined quantifiable goals: factual accuracy greater than 95%, most responses within 2 seconds, and 100% interception of sensitive topics.
Even without a large model API key, the project includes a complete Mock demo mode that can run with zero configuration.
The team wrote automated acceptance scripts covering 63 test cases—including key player queries, factual trap questions, sensitive topic interception, and multi-turn conversations—all of which passed, with one-click regression testing supported.
All Facts (4)
The vast majority of coding for the HoopMind (NBA Chat) project was completed by Claude Code as the primary agent, supplemented by DeepSeek V4 Flash as a sub-agent.
50%UnverifiedThe HoopMind project adopts checklist-driven development, first producing an RFC requirements document with clearly defined quantifiable goals: factual accuracy greater than 95%, most responses within 2 seconds, and 100% interception of sensitive topics.
50%UnverifiedEven without a large model API key, the project includes a complete Mock demo mode that can run with zero configuration.
50%UnverifiedThe team wrote automated acceptance scripts covering 63 test cases—including key player queries, factual trap questions, sensitive topic interception, and multi-turn conversations—all of which passed, with one-click regression testing supported.
50%